Sigma 35mm f/1.4 Arrives. Announces New World Order.

Posted by

OK, I'm beginning to think maybe the Mayans were right. It appears the world is going to transition into something different in 2012. The photography world at least.

About 5 years ago, I wrote a blog post explaining that quality control problems and horrid repair service meant we would probably stop carrying Sigma products entirely. I spent the next several months manning the ramparts and pouring hot oil on the Sigma Fanboys who assaulted the Lensrentals Walls.

Since then, the most amazing thing happened. They got better. The repair center sprouted an efficient and intuitive web page, real people started answering the phones and knew where your stuff was, repair times went from months to weeks, to often days. Quality control seemed to improve, too, except for the large telephoto zooms. Recently they announced ( making announcements - what a concept, Nikon) quality control improvements, redesign of some problematic lenses (OK, they didn't use the word problematic lens, that's me. They just said redesign), and are going to offer the gearheads among us unprecedented ability to fine tune their lenses to our cameras.

So today, we received our first 35mm f/1.4 Sigma lens. The first in their new revamped lineup.  (No, you can't rent it yet. It's going to have to undergo extensive testing at my house over the long weekend. Maybe next week.) I was eager to see it, hoping it was going to be another step forward and hoping to find some signs of what will be adjustable in these new lenses.

As always, this isn't a review, it's my quick first impression after putting the lens through our normal intake tests. I'm not a lens reviewer. Also, as always, my summary comes first, for those of you who have trouble reading more than 150 words without a picture.

This lens kicks butt, takes names, and basically posterizes the manufacturers who make the cameras this lens will fit on. 

For those of you who spend too much time post processing, I'm using 'posterize' as demonstrated below, not the 8-bit jpg way.


copyright Getty Images / Sports Illustrated


The Sigma 35mm f/1.4

Left to right: Canon, Sigma, and Nikon 35mm f/1.4 lenses. Copyright Lensrentals.com by Hostess Joey



OK, you can get out your crayons and color me Fanboy, but this lens is built solidly. It really feels more like a Zeiss 35mm than a Canon or Nikon. Sigma says there's a lot of metal in there, and at 1.4 pounds I believe them. On the other hand, that makes it several ounces heavier than either the Canon or Nikon, but about the same amount lighter than the Zeiss 35mm f/1.4.

It feels solid, too. The manual focusing ring is smooth and accurate, although I found it a bit stiff. Not problem stiff, but certainly not move-it-with-a-fingertip. Autofocus speed was reasonably quick, about on a par with the Canon 35mm. More importantly to me, AF accuracy was good, too, even when we darkened the room down to the point where many Sigma lenses start hunting.

Of course, we went over to the Imatest bench next. As most of you know, I hate testing one copy of a lens. The only thing I hate more than that is testing no copies. So I did the one we have today, knowing that there will be another dozen next week and planning just to keep this data to add to that. But like chicken salad on a hot summer day, this data won't keep.

Of course, I'm going to compare this copy to the average numbers we get for the Canon 35mm f/1.4. I'll mention that this may be the absolute sharpest of the zillion or so of these lenses Sigma turns out. I'll get data for another dozen copies next week and we'll see how it varies. So, just for the sake of argument, let's assume this is the best copy Sigma will ever make. So to make things fair, I'm going to compare it to both the average for all Canon 35mm f/1.4 lenses we have, but also to the very best out of the 100 copies I've tested.


MTFSigma 35mm f/1.4Canon 35 f/1.4 avgCanon 35mm f/1.4 best
Peak MTF 50775650700
Avg MTF 50665555600
Worst Corner MTF 50445325370


As you can see, this copy of the Sigma 35mm f/1.4 was a bit better in the center and clearly better in the corners than the best copy of the Canon 35mm f/1.4 I've ever tested. For the Nikonians amongst us, the Nikon 35mm f/1.4 tested on a D3x resolves almost identically to the Canon.

Stopped down to f/2.0 and f/2.8 respectively, the Sigma responds with 900/770/560 and 1000/870/770 mtf50 values (center/average/corner). The Canon 'best copy' returned 840/740/520 and 1000/840/740 numbers. It's probably worth mentioning the Sigma has slightly less distortion at 1%, too.

The Circuit Board

We didn't do a disassembly today but we did have to take a quick look at the main circuit board, since this will apparently be the first 'totally programmable' lens using the upcoming Sigma Optimization Software.

The PCB board in this lens is quite different from anything we've seen from Sigma before: cleaner with more chips and few other electronic components. Much different.

Also looking like it will be nice is the 9-bladed aperture ring.

I've had about zero chance to actually shoot with it (neither would you if Tyler was standing around screaming about another 50 lenses that need to be tested so they can go on sale for Black Friday) but I'll fix that this weekend. In the meantime, here's a shot of the lovely and talented Lensrentals spokesmodel Kenny, at f/1.4, with a bit of 100% crop in the corner. I believe this lens is going to be fun. And at $899 well worth the price of admission.

As to finding anything signifying the new programmability of theses lenses, I didn't. But if some of the more electronically educated among you want to look up some chip codes, I've got a full size image of the PCB you can look at.

Roger Cicala


November 2012

Addendum: A few real world pictures added, along with the note that 1) the lens is not weather sealed and 2) AF was as fast and accurate as the Canon L primes I usually shoot with. I didn't do side-by-side comparisons, but it was certainly not noticeably different.

Roger Cicala, 2012


Roger Cicala, 2012


Roger Cicala, 2012


Roger Cicala, 2012


Roger Cicala, 2012


The weakness of the lens (every lens has some) does begin to show up in some of these photos: the backround bokeh is not nearly as smooth and buttery as the Nikon or Canon 35 lenses. That's how it is with lenses - the designer trades off one aspect for another. With this lens we have awesome sharpness at an excellent price, and perhaps at the expense of background blur.

For some photographers the background blur will be a deal breaker. For others the price and sharpness will far outweigh that issue. Nice to have choices, isn't it?

126 Responses to “Sigma 35mm f/1.4 Arrives. Announces New World Order.”

Joe said:

No judgement needed. I'm merely reporting my observations, in hope that this will help motivate further investigation, and perhaps lead to some technically plausible explanation, preferably with a lens firmware fixes.

BTW, I've had mis-focused shots with my Sigma 35/1.4 at f/5.0 (high contrast urban scene in good light), so this isn't an optical issue such as lens speed or f-stop. This is an electromechanical focusing consistency issue -- the next shot of the same scene, using the same focusing technique, came out sharp.

In good light, my Sigma mis-focuses perhaps 3% of the time, but clearly stopping down is helping during daytime, because the fraction of mis-focused shots in low light at f/1.4 or f/2.0 is significantly higher (in my experience).

I've been using a variety of Canon AF lenses (fast and slow, primes and zooms, from ultra wide to super telephoto, film and digital) and several EF-mount camera bodies since the original EOS 650 over 25 years ago, and I have not seen this frequency of AF problems. Then again, my Sigma 35/1.4 is optically a delight, vastly superior to my old Canon 35/2, and I agree with Roger's assessment of its optical quality.

Evil Ted said:

I have a copy of the Sigma 35mm F/1.4 for Canon mount and I can also confirm the auto focus problem :(
It's a nice lens and really sharp, but pretty useless if you are shooting stationary objects indoors in low light and it results in soft or out of focus shots.

I'm getting way higher than 3% misses too, more like 25%.
In this respect it reminds me of the Canon 50mm F/1.2L - great when it nails focus, but the problem is it misses too much.

Shawn C. said:

I too am having problems with the Sigma 35mm @ F/1.4 with focusing. It's very inconsistent and tends to front focus most of the time. I had to set it to +17 on my 5D Mark III and this is after days of testing. When it does miss it's still front focusing. I'm wondering if I just got a bad copy of this lens? It does miss 100% of the time with NO AF adjustment. Not having this problem with my other 4 Canon lenses, including the 135mm F/2.

Evan said:

Hi Roger
I have the Sigma 35 1.4 here and I'm having a heck of a time micro adjusting it. The focus seems inconstant. I'm using it with the 5DIII. What are your thoughts on AF accuracy with the Sigma? Your articles about Canon's improved AF accuracy were eye opening but I don't... Wait a minute, I just saw the preceding comments. I guess I'm not alone. I think consistant accurate AF trumps lens sharpness. That's why I'm throwing in the towel with my Zeiss 35mm f/2. It's an amazing lens but my eyes are getting old.

Thomas Pindelski said:

I rented a brand new one (local place in SF, sorry Roger!) and while the lens has the best resolution and color rendering I have seen, it's useless based on my example. It simply cannot focus consistently on my D3x, known to be dead accurate every time with my 85/1.8 at full aperture. I would estimate the hit rate at no better than 50%. Identical exposures, seconds apart, vary between back focus, front focus and dead on. This looks like a design issue with the stepper motor and/or circuitry. The miss rate is so great that the lens simply cannot be trusted in any light - I get the same results in daylight as in poor light at any distance and it will miss even at f/4 by a significant amount on occasion.

I thought mine might be an isolated example but the other recent comments along like lines suggest this is a broader issue.

Roger, how do you test these? Are you using LiveView on your test bench or are you testing without a camera body? I only use regular AF (I'm a street snapper, so LiveView is not useful). Some AF tests from you might be instructive.

LensRentals Employee

Roger Cicala said:

Thomas, all of our published resolution testing is done in live-view on bodies. We don't do any critical autofocus testing other than just shooting around the lab, which means plenty of controlled light, etc.

Thomas Pindelski said:

Thank you, Roger. It's so great optically that I will wait for Sigma to fix the AF issues before buying one.

If it helps I have made an ACR lens correction profile (RAW and DNG only) which can be downloaded here:


I guess it will take Adobe a while to make one.

This works with Lightroom and Photoshop and is especially useful at f/1.4 and f/2 where the Sigma vignettes heavily.

(My site is non-commercial so this is not an attempt to garner clicks) .

LensRentals Employee

Roger Cicala said:

Thank you Thomas! Most useful. I'm going to go grab it myself.

Andy Holdsworth said:

Hi Roger,

I have just received the 35mm and am finding after calibration it works well for subjects 4-8 feet away but the further the subject is thereafter the more it seems back focuses, so I was wondering if you could tell me if this is normal behaviour for this lens. I'm using the same aperture for the testing.

LensRentals Employee

Roger Cicala said:

Andy, I can't say it's normal, but I've heard of some people with similar issues. Sigma can adjust it for you or you can exchange it. In a month or so, when we have the adjustment tool Sigma is promising, you should be able to adjust it yourself but that doesn't help with your copy right now.

Thomas Pindelski said:

The Beta release of Lightroom 4.4 now includes an Adobe profile for the Sigma 35/1.4. I have checked it against mine - referenced in my comment above, and it performs identically.

Bob B. said:

Ok...this is a great lens. I took delivery on one yesterday and will be breaking it in this weekend.
Build is great, fit and finish superb, appears to be VERY sharp, lens hood included and it snaps on REALLY TIGHT...no slop. NICE. Sigma...you almost got everything perfect...except the case. What is with the over-sized toaster cozy???? I put that thing right back in the box and it will never see the light of day again. A nice, simple sexy neoprene pouch would have complemented this piece of Art perfectly...and cost less to manufacture. I know I am quibbling...but that case is truly hysterical!...easily fixed with 3rd party item. Clearly the team that designed that lens had nothing to do with the case. LOL!

Peter said:

I first handled one of these lenses this week and was seriously impressed by its apparent build quality. It is heavily engineered and exudes the look and feeling of being a pro lens. Previously I regarded myself as a kind of Nikon snob and thought of this company being "Stigma" not "Sigma". Handling the lens almost changed my mind.

I rushed home to "Google" the lens to find out what people are saying about it as I now invariably do before investing money in new kit. Having done that, I am distressed to hear so many reports (not only on this forum)of focusing irregularities. And I have to say that until this issue is fixed, buying this lens must be off my agenda.

Apart from a DSLR I also shoot a Leica M8 which of course uses a rangefinder to focus. Focus calibration can be problematic with that camera / lens combination and I have experienced this and had to spend time money and effort to get the camera calibrated properly and make sure specific lenses are working properly in the field. I can only say it is no fun. It is extraordinarily frustrating to have to cope with missing shots randomly because kit is not quite up to the task.

Sad, so near and yet so far.

JP said:

I had bought a Sigma 35mm for my Canon 6D and yes, it front focused. Sent it back to Sigma, who said that they had recalibrated the lens, but that it probably was my camera's fault. This is ridiculous, for I never had any focus problem with any of my 10 other (L and non-L) lenses. The recalibrated version still front focused as much as before. So I went for a refund; it may be a beautiful lens but if it does not focus properly, it is useless for making pictures with it. I have now bought the Canon 35mm 2.0 IS instead, and I am actually happy that I did. It is much sharper in the corners, lighter, smaller, and the IS is remarkably practical. If I need shallow DOF I use a 85 or 135 mm.

Vladimir Kadinski said:

No problems at all with the autofocus of this lens on my D600. It's spot on every time, even in very low light. Detail wide open is amazing. This lens is a stellar performer. The Nikon and Canon alternatives can't touch it. Sharper, better contrast and totally usable center wide open. I returned the Nikon AF-S 35mm f/1.4G for this lens. I think that says a lot.

MA said:

Am I the only one who cannot get this lens to focus on the outer points of the 6D? I have two copies and none works. Center is spot on, outer points are off by the equivalent of +15 MFA. Unsable at f1.4.

Roger you shoot 6D and this lens, any problem with outer points? Thanks

LensRentals Employee

Roger Cicala said:


I haven't had any problems, and we haven't had reported problems with Canon cameras. There are issues with outer focus points on the D800 Nikons, but haven't seen it on Canon.

When you take images using live view focusing are the edges and corners soft? It could be a decentered lens, that can affect off-center focus accuracy.


MA said:

Thanks for replying Roger, if I use live view or center point the corners are fine. Did a quick test for decentering (on a chart) when I got it and it seems also fine. The point is that I have two lenses with same issue...(serial number is only 5 units apart though). In the meanwhile I sent some pics to Sigma and they agree there is frontfocus to be fixed on the laterals. They asked Sigma in Japan if they have a software solution for this. Will let you know when I hear from them. My 6D works just fine with other lenses.

papa2jaja said:

I've had two copies of this lens and have done fairly extensive testing, using exclusively the center AF point on my D800.

I set both copies up using the USB dock, and found that they appear to have both front/back focus issues plus focussing reliability issues.

I'm a hobbyist and not a pro, so it may be all my fault, but for example, if I set my second copy's focus up to be on spot at around 1.2m, I get a severe backfocus at 6m, which does not exist if I leave the lens to have a bit of a front focus at 1.2m.

After approx. 100 test shots, taken in daylight, of the same 6 objects at identical distances, it appears that with everything beyond around 8 meters, sometimes the focus is spot on, sometimes it is off, although the light hasn't changed.

I've read of similar probs from Canon and Sony users as well. Looking at pics on the Internet, I sometimes think that they are not as sharp as this lens can be, or that the focus seems to be in a location where I don't think the photographer actually put it.

Which makes me wonder, is this a general issue with some newer cameras? Or is it an issue of this lens in general?

Has anyone done test shots like me, of the same 6 or 7 objects at always the same distances and seen that some of them were either sometimes sharp and sometimes not (like me), or always sharp?

I love the images this lens produces when it focusses properly. Is there a chance that with a third copy I might have no probs, or is the probability for this low?

Elle Elskamp said:

I had a very inconsistent focus problem with a new lens on both my Canon 7D and 550D.
Brought it back to the shop, who returned it to Sigma.
I am still waiting for an answer 1 1/2 week now, but I have a strange feeling about it.
Not happy at all.

3bc said:

I don't know if I am lucky or if this is a case of people that have problems being the ones to speak up (which is somewhat natural I believe). My copy is brilliant. Right out of the box, first copy. The only time I miss focus is in incredibly low light on my 6D where other bodies would give up looking anyhow, and that is more of a stress test than practical use. The only negative thing about this lens is it makes every other lens I own seem totally inadequate. Sharpness, color rendition, bokeh (which on my copy is creamy and really messy), "it" factor, this lens blows my others away. My other lenses are not the most top notch but generally we'll respected: 50mm 1.4, 85mm 1.8 and 24-105mm f4 (all canon). I just spent 5 days in Europe doing a lot of shooting and of the 1000 images I took maybe 100 weren't with the sigma, and those would be shots I needed all the length I could get on and went to the 24-105.

My typical Lightroom PP involves bumping clarity and saturation up to around 10-13, vibrance to around 13-15 and sharpness to around 45-55 depending on the particular scene on pretty much every shot just to get dullness that RAW files/my kit imparts on my images. I have not had to add any clarity, vibrance or sharpness to any shots I've taken with this lens. Can't recommend highly enough.

Ansel Adams said:

I had several bad focus problem during a wedding few days ago. It gives bad focus at 5.6 too in day light on a 5d mark II using the central focus point. The problem is not the front-back focus i tried to correct with a +15 M/A but the inconsistence you get 90% of the time. Yes it's good at f1.4 very little chromatic aberrations but from 5.6 it's not an impressive lens at all, a "cheaper" canon 85 1.8 is as sharp or maybe sharper.
Reading other comments i am starting to think reviews on internet are all fake or maybe i have only bad luck.

Daniel Hodac said:

I did some extensive testings on 2 copies. Around 300-400 test shots at all distances. Even shooting same subject at the same distance would yield focus inconsistent, shot A would be front focus while shot B was back focus. It drove me crazy.

The lenses seem to "auto" focus on their own. After several attempts and adjustments from Sigma USB Dock to get to +9/+9/+10/+10 on one lens, other lens was at +10/+10/+15/+15. With the above adjustments, both of the lenses produced tack sharp images if they didn't shift focus on their own. Unfortunately, there was only 20%-25% of the chance that the lenses would focus on the focus point.

My works can't accept this kind of inconsistency. It may work perfectly on other cameras but definitely not on Nikon D810.

I love Sigma 18-35mm f1.8 (APS-C) while hate Sigma 35mm f1.4 at the same level.

I returned both copies. Others may have good luck, however, I didn't have that luck.

Eric Bartlett said:

Not happy with this lens at all. Atanything below f2.8 it's useless in studio or in natural light for portraits when shot 8 feet or so away from subject. Eyes are rarely in focus and I have tried even fast shutter speeds and it doesn't help. I wish I had never bought it. Yes, it maybe sharp close up but within a 35mm close up for portraits you have to be careful with distortion. A BIG disappointment and I don't get it when it comes to good reviews. My Nikon 85mm f1.4 and 70-200 f2.8 knocks the socks off this lens. You get what you pay for I guess. IO will be sticking to brand from now on despite reviews on Sigma.

Eric Bartlett said:

Found out my Sigma f1.4 35mm lens was way off on calibration which is something I overlooked. Added a +20 adjustment using SpyderLenscal tool and voila! Tested this afternoon in an outdoor model shoot and it is working like a charm. Maybe others in this forum who find this same problem need to also calibrate the lens. Apparently, calibration is often overlooked by photographers.

Joe said:

@ Roger Cicala:
Read your test with interest, since I myself had severe problems with the new ART-line of Sigma and wanted to jump in with some observations, thoughts and remarks.
- I got the 18-35mm/1.8 ART: had it shipped to Sigma service 4 (!) times, once even including my Nikon (that previously was sensor- and autofocus-tested by Nikon's service to be on the sure side): even though they claimed to have tested and adjusted the lens, autofocus on anything more distant than 2 meters was still WAY off all the time. Even after the calibration, they have claimed to have done ON MY CAMERA! So blurry, that you even could see that right on the camera's LCD-display without zooming into the photo. Sigma even sent me a new lens in exchange the fourth time: marginally better, but still way off. Cross-testing with other cameras from my dealer: it WAS the lens, not the camera. After some web research, this seems to be a quite common problem: either you are lucky and got a good copy, or you've just lost...
- the adjustment tool (USB dock), that I have, did not help: yes, you can adjust the lens. But: even if you got infinity-focus adjusted (next adjustment point is 50cm), you still will be presented with 2 problems: objects between 2m up to around 30m will still not be "focus-eable". So anything between the two adjustment-points will stay problematic. And: you still will have problems with focus inconsistency: maybe, one time focus is spot on, but the next 10 shots will be non-useable. Tested it in a real-world situation at a well-lit dance-contest: with single spot focus, NOT ONE of 500 pictures was sharp (and I am not talking about motion blur, but severe mis-focusing)!
- You said, you test with Live-view focus. IMHO, this will give you an idea, what a lens might be capable of, but not, how the lens behaves in real-world situations. The big problem, the ART-series seems to have is with "Phase-Autofocus" - which I think is the autofocus setting used in 95% of all cases. I need fast and reliable autofocus, and I can't wait, until live-view autofocus has done its task. Would anyone doing street photography, documentary style or just normal everyday "shoot what I find interesting" use live-view autofocus - I would say "no", they all use "normal" phase-autofocus. From lenses in that price-category, I expect reliability. Not a single one of my Nikon lenses, and not my Sigma 8-16mm and 80-400mm have the slightest problem with phase autofocus. They are spot on. Only the 18-35mm ART (and, as I read but cannot prove myself, other lenses from the ART series as well). May I suggest, you include "normal" autofocus in your testing?

Let me add a final remark: this post is not about bashing Sigma. I just wanted to exchange some thoughts and observations. Besides guys like you testing equipment, user-experiences are the only way for us end-users to get an idea about manufacturer's quality control and possible problems of a lens. Thanks for providing that opportunity to express our observations in your blog.

Leave a Reply